Science, 159(3810): 56-63, January 5, 1968
.
The Matthew Effect m Science
0
The reward and communication systems
of science are considered.
Robert K. Mcrton
,\
This paper develops a conception of
ways
in which certain
psychosociai
\
., processes affect the allocation
of rewards
to
scientists
for
their
contribu\
tjons- an allocation which in turn affwts the flow of ideas and findings
th&gh
the communication
networks
of s&,nce. The conception
is based
upon ah, analysis of the compotite
of
experienc; ,,reported in Harriet Zuckerman’s intemiews with Nobel laureates
in the United @ates (1) and upon data
drawn from the diaries, letters, notebooks. scientific &pen, and biographies
of other scientists.
image and the pubiic image of scientists arc iargeiy shaped by the cornmunaily validating testimony of sign&
cant others that they have variously
lived up to the exacting institutional
requirements of their roies.
A number of workers, in empiricai
studies, have investigated various aspects of the reward system of science
as thus conceived. Gtaser (3) has found.
for example, that some degree of rcco&ion
is required to stabiiizc the
careen of scientists. In a case study
Crane (4) used the quantity of publication (apart from quality) as a measure
of scientific productivity and found that
highly productive scientists at a major
The Reward System and bceaparats
university gained recognition
more ofof the Forty-First chair”
ten than equally productive
scientists
’
‘at a lesser university. Hagstrom (5) has
We might best begin with s&me gendeveloped and partly tested the hypotherai observations on the reward system
esis that matcriai rewards in science
in science, basing these on eariier ibexfunction primarily to reinforce the opreticai formulations
and empirical ‘ineration of a reward system in which
vestigations. Some time ago (2) it wad
the primary reward of recognition
for
noted that graded rewards in the realm ” , scientific contributions
is exchanged for
of science arc distributed principally in
+cess to scientific information.
Storer
the coin of recognition
accorded
re(6) ,has analyzed the ambivalence
of
search by feUow-scientists. This recogthe @cntist’s response to recognition
nition is stratified for varying grades
“as a qase in which the norm of disOf scientific accompiishment,
as judged
intereste+ss
operates to make scienhv. the scientist’s peen. Both the selftists deny‘ ,thc vaiue to them of influence and kvthority in science.” ZuckThe author is Giddings Professor of Sociology
at Columbia University, IJew York 10027. Thi8
erman (7) and S,he Coics (8) have found
article is brKd on a papw read before the
that scientists ~$0 receive recognition
Amctican Sociolo@icrJ Association in San Francisco. August 1967.
for research done early in their ca-
rcers are more productive later on fhnn
those who do not. And the Cola &WC
ato found that. at least in the B
of contemporary American phys& the
reward system operates iargciy in a*
cord with institutional
valua
of the
s&linasmuch as quality of d,
is more often and more su-9
rewarded than mere quantity.
**
In science as in other ins&u&d
f&n&
a special problem in &8 w&inlp of the reward system tums up
when individuals or organizations
take
on the job of gauging and suitably
mVaKiing lofty performance
on behalf
of a large community. TIiw M uitim
accolade in ZOth*eattay
sciatc&
the Nobel prize, is often- asshed
to
mark off its recipients from ail the
other scbtists
of the time. Yet. tbis
assumption
is at odds with the- wellknown fact that a good number of
scientists who have not recdved the
prize and wiil not receive it ham contributed as much to the advrrrrccnrtnl
of science as some of the iedpients.
or more. This can be desctibed a the
phenomenon
of “the 4lst ChaiP The
derivation of this tag is citar enough.
The French Academy,
it M
be rcmembered. decided early that oniy a
cohort of 40 couid qualify as m-0
hen and so emerge as immo&
This
limitation of numbers ma& inevitable,
of course, the exciusi6n through the
centuries of many talented individrrallr
who have won their own immortality.
The familiar list of occupants Of this
41st chair inciudes Descartes, Pd.
Moli&rt, Bayle, Rousseau, Saint&non,
Didcrot, Stendahl, FIaubert, 204
and
Proust (9).
What holds for the French Academy
holds in varying degree for every bther
institution designed to identify and reward talent. i’n all of them there arc
occupants of the 4lst chair, mea outside the Academy having at least the
same order of talent as those inside it.
In part, this circumstance
results from
emrs of judgment that lead to inclusion of the less talented at the expense
of the more talented. History serves
1
Copyright @ 1968 by the
Amencan Association for the Advancement of Science
Reprinted from
SCIENCE
Januuy 5, 1968. Vol. 159, No. 3810. pages 56-63
Y
.
$.
as an appellate court, ready to reverse
Jacques Loeb, W. M. Bayiiss, E. H.
the judgments
of the lower courts,
Starling, G. N. Lewis, 0. T. Avery,
which are limited by the myopia of and Selig Hecht, to say nothing of the
long list of still-living uncrowned Nobel
contemporaneity.
But in greater part,
laureates (IO).
the phenomenon
of the 41st chair is
In the stratification
system of honor
an artifact of having a fixed number
in science, there may also be a “ratchet
of places available at the summit of
effect” (II) operating
in the careers
recognition. Moreover, when a particular generation is rich in achievements
of scientists such that, once having
achieved a particular
degree of emi-’
of a high order, it follows from the
rule of fixed numbers that some men
nence, they do not later fall much below that level (although
they .may be
whose accomplishments
rank as high
as those actually given the award will
outdistanced by newcomers and so sufbe excluded from the honorific ranks. ’ fer a relarive decline in prestige). Once
a Nobel laureate, always a Nobel lauIndeed, their accomplishments
somereate. Yet the reward system based on
times far outrank those which, in a time
recognition for work accomplished tends
of less creativity,
proved enough to
to induce continued effort, which serves
qualify men for .&is high order of
both to validate the judgment that the
rccognitioa
scientist has’ unusual capacities and to
The Nobel prize retains its luster betestify that these capacities have concause errors of the first kind-where
tinuing potential. What appears from
scientic
work of dubious or inferior
below to be the summit becomes, in
worth has been mistakenly
honored
the experience
-are uncommonly
few. Yet limitations
of those who have
reached it, only another way station.
of the second kind cannot be avoided.
The scientist’s ,peers and other ass00
The small number of awards means
ciates regard each of his scientific
that, particularly
in times of great
achievements
as only the prelude to
scientific advance, there will be many
new and greater achievements. Such sooccupants of the 41st chair (and, since
cial pressures do not often permit those
the terms governing the award of the
who have climbed the rugged mouno
prize do not provide for Posthumous
tains of scientific achievement
to rerecognition,
permanent
occupants
of
main content. It is not necessarily the
that chair). This gap in the award of
fact that their own Faustian aspirations
the ultimate prize is only partly filfed
are ever escalating that keeps eminent
by other awards for scientific accomscientists at work More and more is
plishrnent since these do not carry the
same prestige either inside the scientific
expected of them, and this creates its
community or outside it. Furthermore,
own measure of motivation and stress.
Less often than might be imagined is
i*what has been noted about the artifact
there repose at the top in science
of fixed numbers producing occupants
of the 41st chair in the case of the
(see 12).
The recognition
accorded scientific
Nobei prize holds in principle for other
achievement by the scientist’s peers is
awards providing less prestige (though
a reward in the strict sense identified
sometimes, nowadays, more cash).
by Parsons (13). As we shall see, such
Scientists reflecting on the stratificarecognition
can be converted into an
tion of honor and esteem in the world
instrumental
asset as enlarged facilities
o&&&e
know all this; the Nobel
are made available to the honored scien*‘laureates themselves know and emphatist for further work. Without deiibersize it, and the members of the Swcdish Royal Academy of Science and the
ate intent on the part of any group.
the reward system thus influences the
Royal Caroline Institute who face the
“class structure” of science by providunenviable
task of making the tial
ing a stratied
distribution
of chances,
decisions know it. The latter testify
to the phenomenon
of the 41st chair
among scientists,
for enlarging
their
whenever they allude to work of “prizerole as investigators. The process provides differential
access to the means
winning calibre” which, under the conditions of+ the scarcity of prizes, could
of scientific production.
This becomes
not be given the award. And so it is
alf the more important in the current
*historical shift from little science to big
that, in the case of the Nobel prize, occupants of the 41st chair comprise an
science, with its expensive and often
illustrious company that includes such
centralized
equipment
needed for renames as Josiah Willard Gibbs, Mendesearch. There is thus a continuing interplay between the status system, based
Ieev, W. B. Cannon, H. Quincke, J.
Barcroft, F. d’H&elle, H. De Vries,
on honor and esteem, and the class
2
system,
based . on differential
iifechances, which locates scientists in differing positions within the opportunity
structure of science (14).
The Matthew
Effectin the
RewardSystem
The social structure of science provides the context for this inquiry into
a complex psychosocial
process that
affects both the reward system and
the communication
system of science.
We start by noting a. theme that runs
through the interviews with the Nobel
laureates. They repeatedly observe that
eminent scientists get dispmportionately great credit for their contributions
to science while reiativeiy unknown
scientists tend to get disproportionately
little credit for comparable
contributions. As one laureate in physics put
it (25): “The world is peculiar in this
matter of how it gives credit. It tends
to give the credit to [already] famous
people?
As we examine the experiences reported by eminent scientists we find
that this pattern of recognition, skewed
in favor of the established scientist, appears principally
(i) in cases of co10
laboration
and (ii) in cases of independcnt multiple discoveries made by
scientists of distinctly
different rank
(Ia)
1; papers coauthored by men of decidedly
unequal
reputation,
another
laureate in physics reports, “the man
who’s best known gets more credit, an
inordinate
amount of CrcdW
In the
words of a laureate
in chemistry:
“When people see my name on a paper,
they are apt to remember it and not to
remember the other names.” And -‘a
laureate in physiology
and medicine
describes his awn pattern of response
to jointly authored-papers.
,
You usually notice the name that you’re
familiar with. Even if it’s fast. it wilf be the
one that sticks. In some cases, all the
names are unfamiliar to you, and they’re
virtually anonymous. But what you note
is the acknowledgement at the end of the
paper to the senior person for his “advice
and encouragement.” So you wiil say:
“This came out of Greene’s lab, or se
and-so’s lab.” You remember that, rather
than the long list of authors,
Almost as though he had been listening to this . account, another laureate
in medicine explains why he will often
not put his name on the published report of a collaborative piece of work:
“People are more or less tempted to
.
say: ‘Oh yes, so-and-so is working on
such-and-such
in C’s laboratory.
It’s
C’s idea.’ I try to cut that down.” Still
another laureate in medicine alludes to
this pattern and goes on to observe
how it might prejudice the career of
the junior investigator:
If someone
is being conside=d for a job
by people who have not had much experience with him, if he hw published
only together with some known name+
weii, it detracts. It naturally makes people
ask: “How much is really his own contribution. how much [the senior author’s].
How will he work out once he goes out
of that laboratory?”
.
Under certain conditions this adverse
effect on recognition of the junior author of papers written in collaboration
with prominent scientists can apparent.- ly be countered and even converted
into an asset. Shouid the younger scientist .move ahead to do autonomous and
significant work, this work rerroactivciy
affects the appraisals of his role in earlier collaboration. In the words of the
laureate in medicine who referred to
the virtual anonymity
of junior authors of coauthored papets: “People who
have been identified with such joint
work and who then go on to do good
work later on, [do] get the -proper
amount of recognition.” Indeed, as another laureate implies, this retroactive
judgment may actuaily heighten recognition for later accomplishments:
“The
junior person is sometimes iost sight of,
but only temporariiy if he continues.
In many cases, he actually gains in acceptance of his work and in generai
acceptance, by having once had such
association.” Awareness of this pattern
of retroactive recognition may account
in part for the preference. described by
another laureate of some “young feilows [,who] feel that to have a betterknown name on the paper wiil be of
help to them.” But this is an expressive
as weii as a merely instrumental preference, as we see also in the pride with
which laureates themselves speak of
having worked, say, with Fermi, G. N.
Lewis, Meycrhof, or Niels Bohr.
So much for the misallocation
of
credit in this reward system in the case
of collaborative work. Such misaiiocation also occurs in the case of independent multiple discoveries. When approximately the same ideas or findings
are independently
communicated
by a
scientist of great repute and by one
not yet widely known, it is the first,
we are told, who ordinarily receives
prime recognition.
An approximation
.
to this pattern
is reported
reate who observes:
by a iau-
It does happen that two men have the
same idea and one becomes better known
for it. F-+
who had the idea, went circling round to try to get an experiment
for. .
. Nobody wouid do it and so it
was forgotten, practically. Finally, L
and c‘,, did it, became famous.
and Band got the Nobel Prize. . . . If things
had gone just a little differently; if somcbody had been willing to try the txpcriment when E
suggested it, they probably couid have published it jointly and he
would have &en a famous man. As it is,
he’s a footnote.
l
The workings of this process at the
expense of the young scientist and to
the benefit of the famous one is remarkably summarized in the life history of a laureate in physics, who has
experienced
both phases at diRerent
times in his career.
When you’re not recognized, he recalls,
it’s a little bit irritating to have somebody
come along and figure out the obvious
which you’ve also figured out, and everybody gives him credit just because he’s a
famous physicist or a famous man in his
field.
Here he is viewing the case he reof one who
had this happen to him before he had
become famous. The conversation takes
a new turn as he notes that his own
position has greatly changed. Shifting
from the perspective of his earlier days,
when he felt victimized by the pattern.
to the perspective of his present high
status, he goes on to say:
ports from the perspective
This often happens, and I’m probably
getting credit now, if I don’t watch myself,
for things other people figured out. Because I’m notorious and when I say it,
people say: “Well, he’s the one that
thought this out.” Well. I may just be
saying things that other people have
thought out before.
In the end, then, a sort of rough-hewn
justice has been done by the compounding of two compensating injustices. His
earlier ~accomplishments have been underestimated;
his later ones, overestimated (I 7).
This complex pattern of the misaiiocation of credit for scientific work
must quite evidently be described as
“the Matthew effect,” for, as will be
remembered, the Gospel According to
St. Matthew puts it this way:
For unto every one that hath shall be
given, and he shail have abundance: but
from him that hath not shall be taken
away even that which he hath.
Put in less stately language, the Mat3
thew effect consists in the accruing of
greater increments of recognition for
particuiar scientific contributions to scientists of considerable repute and the
withholding of such recognition from
scientists who have not yet made their
mark. Nobel laureates provide presumptive evidence of the effect. since they
testify to its occurrence, not as victims
-which
might make their testimony
suspect-but
as unwitting beneficiaries.
The laureates and other eminent
men of science are sufficiently aware
of this aspect of the Matthew effect
to make apeciai efforts to counteract
it. At the extreme, they sometimes rcfuse to coauthor a paper reporting rem
search on which they have coilabootcd
in order not to diminish the-.ra%gnition accorded their less-well-known associates. An& as Harriet Zuckcrman
has found (28), they tend to give fint
place in jointly authored papers to one
of their collaborators. She discovered,
moreover, that the iaureates who have
attained eminence before receiving the
Nobei prize begin to transfer fintauthorship
to associates earlier than
less eminent
laureates-to-be
do, and
that both sets of laureate-he
previousiy eminent and not-so-eminentgreatly increase this practice a@r receiving the prize. Yet the latter effort
is probably more expressive of the laureates’ good intentions than it is &cctive in redressing the imbalance
of
credit attributable to the Matthew effect. As the laureate quoted by Harriet Zuckerman
acknowledges:
“If I
publish my name first, then everyone
thinks
the others
are just technicians. . . . If my name is last, people
will credit me anyway for the whole
thing, so I want the others to have a bit
more glory.”
The problem of achieving a public
identity in science may be deepened
by the great increase in the number
of papers with several authors (2, chap.
3; 19; 20, p. 87) in which the -role of
young collaborators
becomes obscured
by the brilliance that surrounds their
illustrious co-authors. So great is this
problem that we are tempted to turn
again to the Scriptures to designate
the status-enhancement
and status-suppression components
of the Matthew
effect. We can describe it as “the Ecclesiasticus component,” from the famil“Let us now praise faiar injunction
book
mous men,” in the noncanonical
of that name.
It will surely have been noted that
the laureates perceive the Matthew tf-
.
feet primarily
as a problem in the
just allocation
of credit for scientific
accomplishment.
They see it largely in
terms of its action in enhancing rank
or suppressing
recognition.
They see
it as leading to an unintended
double
injustice, in which unknown scientists
are unjustifiably victimized and famous
ones, unjustifiably
benefited. In short,
they see the Matthew effect in terms
of a basic inequity
in the reward
system that affects the careers of individual scientists. But it has other implications for the development of science, and we must shift our angle of
theoreticai vision in order to identify
them.
and intensity with the exponential increase (20, chaps. 1 and 2; 26) in the
volume of scientific publications, which
makes it increasingly difficult for scientists to keep UP with work in their
field. Bentley Glass (27) is only one
among many to conciude that “perStudies gf the reading practices of
haps no problem facing the individual
scientists indicate that the suggested
scientist today is more defeating than
possibility-“Nobody
reads
it”“is
the effort to cope with the flood of
something less than sheer hyperbole. It
published scientific research, even with=
has been found, for example, that oniy
in one’s own narrow specialty.” Studabout half of 1 percent of the articles
ies of the communication
behavior of
published in journals of chemistry are
scientists (28) have shown that, conread by any one chemist (22). And
fronted with the growing task of idea&
much the same pattern has been found
fying significant work pubiished in their
to hold in psychology (23, p. 9):
field, scientists search for cues to what
they should attend to. One such a
js’
The data on current readership (i.e., within
the professional reputation 6f the” aua couple [of] months after distribution of
The Matthew Effect
the journal) suggested that about one-half
thors. The problem of locating the perof the research reports in “co&* joumais
tinent research literature and the prob,
in the Commtutication System
wiil be read [or skimmed) by 1% or less
lem of authors’ wanting their work to
of a random sample of psychologists. At
We now look at the same social
be noticed and used are symmetricai:
the highest end of the current readership
distribution, no research report is likeiy to
phenomena from another perspectivethe vastly increased bulk of publicabe read by more than about 7% of such
not from the standpoint of individual
tion stiffens the competition
he-n
a safnpie.
careers and the workings of the repapers for such notice. The American
ward system but from the standpoint
Psychological
Association
study (23,
Several of the Coles’s findings (24)
of science conceived of as a system
pp. 252, 254; 29) found that from 15 to
bear tangentially
on the hypothesis
of communication.
This perspective
23 percent of the psychologist-readers’
about the communication
function of
yields a further set of inferences.
It
behaviors
in selecting articles were
the Matthew effect. The evidence is
leads us to propose the hypothesis that
based on &he identity of the authors.
tangential rather than centrai to the
a scientific contribution will have greatThe workings of the Matthew eflect
hypothesis since their data deal with
er visibiIity in the community of scienin the communication
system require
the degree of visibility of the entrre
tists when it is introduced by a scienus to draw out and emphasize certain
corpus of each physicist’s work in the
tist of high rank than when it is introimplications about the character of scinational community of physicists rather
duced by one who has not yet made
ence. They remind us that science is
than with the visibility of particular
his mark. In other words, considered
not composed of a series of priv>te
papers within it. Still, in gross terms,
in its implications for the reward sysexperiences of discovery by many. S&
their findings are at least consistent with
tem, the Matthew effect is dysfunctionentists, as sometimes seems to be asthe hypothesis. The h&her
the rank
_
al for the careers of individual sciensumed in inquiries centered exclusively
of physicists (as measured by the prestists who are penaiized in the early
on the psychological processes involved
tige of the awards they have received
stages of their development,
but conin discovery. Science is public, not
for scientific work), the higher their
sidered in its implications for the comprivate. True, the making of a disvisibility in the national
community
munication system, the Matthew effect.
covery is a complex personal experi=+
of physicists. Nobel Iaureates have a
in cases of collaboration
and multiple
ence. And since the making of the disvisibility score (25) of 85; other memdiscoveries, may operate to heighten
covery necessarily precedes its fate, the
bers of the National Academy of Scithe visibility of new scientific communature of the experience is the same
ences, a score of 72; recipients
of
nications. This is not the first instance
whether the discovery temporarily fails
awards *having less prestige, a score
of a social pattern’s being functional
to become part of the socially shared
of 38; and physicists who have refor certain *aspects of a social system
culture of science or quickly becomes
ceived no awards, a visibility score of
and dysfunctional
for certain individa functionally
significant part Of that
17. The Coles also find (24) that the
uals within that system. That, indeed,
culture. But, for science to be advisibility of physicists producing work
is a principal
theme
of classical
vanced, it is not enough that fnribful
of high quality is heightened by their
tragedy (21).
ideas be originated or new experiments
attaining honorific awards more presSeveral laureates have sensed this sodeveloped or new problems formulated
tigious than those they have previouscial function of the Matthew effect,
or new methods instituted. The innovaly received.
Further
investigation
is
Speaking of the diiemma
that contions must be effectively communicated
needed to discover whether these same
fronts the famous man of science who
to others. That, after all, is what we
patterns hold for differences in the visdirects the work of a junior associate,
mean by a contribution
to scienceibility (as measured by readership) of
one of them observes:
wmething given to the common fund
individual
papers published by s&nof knowledge. In the end, then, science
tists of differing rank.
It raises the question of what you are to
is a socially shared and socially vaiiThere
is
reason
to
assume
that
the
, do. You have a studentt shouid you put
dated body of knowledge. For the de=
communication
function
of
the
Matyour name on that paper or not? You’ve
velopment of science, only work that
thew effect is increasing in frequency
contributed to it, but is it better that you
shouldn’t or should? There are two sides
to it. If you don’t [and here comes the
decisive point on visibility], if you don’t,
there’s the possibility that the paper may
go quite unrecognized. Nobody reads it. If
you do, it might be recognized, but then
the student doesn’t get enough credit.
js effectiveiy perceived and utilized by
other scientists, then and there, matters.
In investigating
the processes that
shape the development of science, it is
therefore important to consider the social mechanisms that curb or facilitate
the incorporation
of would-be contributions into the domain of science.
Looking at the Matthew effect from
this perspective, we have noted the distinct possibility that contributions made
by scientists of considerable standing
are the most likely to enter promptly
and widely
into the communication
networks df science, and so to accelerate its development.
.
The Matthew
Effect and
the Functions of Redundancy
Construed in this way, the Matthew
effect links up with my previous studies of the functions of redundancy in
science (30). When similar discoveries
are made by two or more scientists working independently
(“multiple discoveries”), the probability that they will be
promply incorporated
into the current body of scientific knowledge is increased. The more often .a discovery
has been made independently, the better are its prospects of being identified
and used. If one published version of
the discovery is obscured by “noise”
in the communication
system of science, then another vemion may become visible. This leaves us with an
unresolved question: How can one estimate what amount of redundancy
in
independent efforts to solve a scientific
problem wiil give maximum probability
of solution without entailing so much
replication of effort that the last increments wiil not appreciably increase the
probability? (See 31.)
In examining
the functions of the
Matthew effect for communication
in
science. we can now refine this conception further. It is not only the number
of times a discovery has been independently made and published that affects its visibility but also the standing,
within the stratification system of science, of the scientists who have made
it. To put the matter with undue simplicity, a single discovery introduced
by a scientist of established reputation
may have as good a chance of achieving
high visibility as a multiple discovery
variously introduced by several scientists no one of whom has yet achieved
a substantial reputation. Although the
general idea is, at this writing, tentative, it does have the not inconsiderable
man of science serves distinctive functions. It makes a diEerence, and.aftcrr
a decisive difference, for the advancement of science whether a composite
of ideas and findings is heavily concentrated in the work of one man or
one research group or is thinly dispersed among a great number of m
and organizations. Such a wmposi~
tends to take on a structure sooner in
the first instance than in the second.
It required Al Freud, for instanto
focus the attention of many psychoi+
gisti upon a tide array of idela which,
as has been shown elsewhere (30), had
in large part also been hit upon by
various other scientists. Such focJidna
may turn out to be a distinctiVe*fUSiCtian of eminent men of science (36. _
A Freud, a Fermi, and a Q&r&
play a charismatic role in t SC&U
They excite intelIectual
en&&am
among others who ascribe cm
q&i&
to them. Not only do thw
themselves achieve excellence, they brvr,
the capacity for evoking exccllEnm in
othen. In the compelling phrase of W
laureate, they provide a “bright am&
ancc.” It Is not so much tha8 thr-rr
great men of science pass on’their te&
niqum methods, information, and theory to novices working with thsoa,
More consequentially, they cm
to
their associates the norms and vahaes
that govern significant research. oiben
in their later years, or after their dti
this personal influence becomes e
ized, in the fashion described by Mm
Weber for other fields of hm
a&~ity. Charisma becomes institutionriized,
in the form of schools of thought and
research establishments.
The role of outstanding men of sciencc in influencing younger as34ktes
is repeatedly emphasized in the in-views with laureates. Almost to a m
they lay great emphasis on the importance of problem-finding, not onfy prob,
lem-solving. They uniformly exprus the
that what rm@tcrs
stmng conviction
most in their work is a d-g
sense of taste, of judgment, in tig
upon problems that *areof fundamtntai
importance. And, typically, they rcpart
that they acquired this sense f
the
significant problem during their yof training in evocative environmmts.
Reflecting on his years as a novice in
the laboratory of a chemist of the first
Social and Psychologicai Bases
rank, one laureate reports that he “led
me to look for important things, whenof the Matthew Ukct
ever possible, rather th& to work on
Even when some of his contributions
endless detail or to work just to imhave been independently made by an
prove accuracy rather than malting a
aggregate of other scientists, the great
basic new contribution.”
Another de-
virtue of lending
itself to approximate
citation indexes
to find whether in multiple discoveries
by scientists of markedly unequal rank
it is indeed the case that work published by the scientists of higher rank
is the more promptly and more widely
cited (32). To the extent that it is, the
findings will shed some light on the unplanned consequences of the strat@ation system for the development of science. Interviews with working scientists
about their reading practices can also
supply data bearing on the hypothesis.
So much for the link between the
Matthew effect and the functions of
multiple discoveries in increasing both
the probability and the speed of diffusion of significant new contributions
to - science. The Matthew effect aIs0
links up with the finding, reported clsewhere (33). that great talents in science
are .typicaily involved in many multiple
discoveries. This statement holds for
Galileo and Newton: for Faraday and
CIerk Maxwell; for Hooke, Cavendish,
and Stensen: for Gauss and Laplace; for
Lavoisier, Priestley, and Scheele: and
for most Nobel laureates. It holds, in
short, for all those whose place in the
pantheon of science is largely assured.
however much they may differ in the
scale of their total accomplishment.
The greatness of these scientists rests
in their having individually contributed
a body of ideas, methods, and results
which, in the case of multiple discoveries, has also been contributed by a
sizable aggregate of lem talented men.
For example, we have found that Kelvin had a part in 32 or more multiple
discoveries, and that it took 30 other
men to contribute
what Kelvin himseif contributed.
By-examining the interviews with the
laureates, we can now detect some
underlying
,psychosocial
mechanisms
that make for the greater visibility of
contributions
reported by scientists of
established reputation. This greater visibility is not merely the result of a
halo effect such that their personal
prestige nibs off on their separate contributions.
Rather, certain aspects of
their socialization, their scheme of values, and their social character account
in part for the visibility of their work.
test. One can examine
scribes his socialization in a European
laboratory
as “my first real contact
with first-rate creative minds at the
high point of their power. I acquired
a certain expansion of taste. It was
a matter of taste and attitude and,
to a certain extent, real self-confidence.
I learned that it was just as difficult
to do an unimportant experiment, often
more difficult, than an important one.”
There is one rough measure of the
extent to which the laureates
were
trained and influenced in particularly
creative
research
environments-the
number of laureates each worked under in eariier years. Of 55 American
Laureates, 34 worked in some capacity,
as young men, under a total of 46
Nobel prize winners (35). But appareatly it is not only the experience of the
laureates (and, presumably,
other outstanding men of science) in these environments that accounts for their tendency to focus on significant problems
and so to affect the communication
function of the Matthew effect. Certain aspects of their character also play
a part. With few exceptions, these are
men of exceptional ego strength. Their
self-assurance
finds varied expression
within the context of science as a social institution. That institution,
as we
know, includes a norm calling for autonomous and critical judgment
about
one’s own work and the work of others.
With their own tendencies reinforced
by such norms, the laureates exhibit
a distinct self-confidence (which, at the
extreme, can be loosely described as
attractive arrogance). They exhibit a
great capacity to tolerate frustration
in their work, absorbing repeated failutfs - without
manifest
psychological
damage. One laureate alluded to this
capacity while taking note of the value
of psychological support by colleagues:
Rpsar& is a rough game. You may work
for months, or even a few years, and seemingly you are getting nowhere. It gets
pretty dark at times. Then, all of a sudden,
you get a break. Jt’s good to have somcbody around to give a bit of encouragtmeat when it’s needed.
Though attentive to the cues provided by the work of others in their
field, the Nobelists are self-directed
men, moving confidently into new fields
of inquiry once they are persuaded that
a previous one has been substantially
mined. In these activities they display
a high degree of venturesome fortitude.
They are prepared to tackle important
though difficult problems rather than
settle for easy and secure ones. Thus, a
laureate recalls having been given, early
in his career, “a problem about which
there was no risk. All I had to do
was to analyze [the chemicai composition of certain materials]. You could
not fail because the method was well
established. But I knew I was going
to work on the tinstead and the
whole thing would have to be created
because nothing was known about it.”
He then went on to make one of his
prime contribution;
in the more risky
field of investigation (36).
This marked ego strength links up
with these scientists’ selection of important problems in at least two ways.
Being convinced that they will recognize an important problem when they
encounter it, they are willing to bide
their time and not settle too soon for
a prolonged commitment to a comparatively unimportant
one. Their capacity
for delayed gratification, coupled with
leads to a conviction
self-assurance,
that an important problem wiil come
along in due course and that. when it
does, their acquired sense of taste will
enable them to recognize it and handle
it. As we have seen, this attitude has
been reinforced by their early experience in creative environments.
There.
association with eminent scientists has
demonstrated
to the talented novice.
as didactic teaching never could, that
he can set his sights high and still
cope with the problem he chooses.
Emulation
is reinforced by observing
successful, though often delayed, outof the
the idiom
comes.
Indeed,
laureates reflects this orientation. They
like to speak of the big problems
and the fundamental
ones, the important
problems
and the beautiful
ones. These they distinguish from the
pedestrian work in which they engage
while waiting for the next big problem to come their way. As a result
of all this, their papers are apt to
have the kind of scientific significance
that makes an impact, and other scientists tend to single out their papers
for special attention.
The character structure of these ieading scientists may contribute
to the
communication
aspect of the Matthew
effect in still another way, which has
to do with their mode of presenting
their scientific work. Confident in their
powers of discriminating
judgment-a
confidence that has been confirmed by
the responses of others to their previous
work-they
tend, in their exposition. to
emphasize
and, develop the central
ideas and findings and to play down
peripheral ones. This server, to highlight the significance of their contribu6
tions, raising them out of the stream
of publications by scientists having less
socially-validated self-esteem, who more
often employ routine exposition.
Finally, this character structure and
an acquired set of high standards often
lead these outstanding scientists to discriminate between work that is worth
publishing
and that which, in their
candid judgment, is best left unpublished though it could easiiy find its
way into print. The laureates and other
scientists of stature often report scrapping research papers that simply did
not measure up to their own demanding standards or to those of their coilegues (37), Seymour Benzcr, for example, tells of how ahe was saved from
going “down the biochemical drain”:
“Delbriick saved me, -when he wrote
to my wife to tell me to stop writing so
many papers. And I did stop” (38).
And a referee’s incisive report on a
manuscript sent to a journal of physics
asserts a relevant consequence of a scientist’s faiiure to exercise rigorous
judgment in deciding whether to publish or not to publish: “If Cwould
write fewer papers, more peopie would
read them.” Outstanding scientists tend
to develop an immunity to insanabilc
scribendi
ctacoethes (the itch to publish) ‘(39). Since they prefer their published work to be significant and fruitful rather than merely extensive, their
contributions
are apt to matter. This
in turn reinforces the expectations of
their feilw scientists that what these
eminent scientists publish (at least during their most productive period) will
be worth close attention (40). Once
again this makes for operation of the
Matthew effect, as scientists focus on
the output of men whose outstanding
positions in science *have been socially
validated by judgments of the average
quality of their past work. And the
more closely the other scientists attend
to this work, the more they are likely
to learn from it and the more discriminating their response is apt to be (42).
For all these reasons, cognitive material presented by an outstanding scientist may have greater stimuius vaiue
than roughly the same kind of material presented by an obscure one-a
principle which provides a sociopsychological basis for the communication
function of the Matthew effect. This
principle represents a special appiication of the self-fulfilling
prophecy
(42). somewhat as follows: Fermi or
Pauling or G, N. Lewis or l
Weisskopf
see fit to report this in print and so
it is apt to be important (since, with
,
some consistency, they have made important contributions in the past); since
it is probably important, it should be
read with special care; and the more
attention one gives it, the more one
is apt to get out of it. This becomes
a self-confirming
process, making for
the greater evocative effect of publications by eminent men of science (until
that time, of course, when their image
among their fellow scientists is one of
men who have seen their best daysan image, incidentally, that corresponds
with the self-image of certain laureates
who find themselves outpaccd by onrushing generations of new men).
Like other self-fulfilling prophecies,
this one becomes dysfunctionai
under
certain conditions. For although eminent scientists may be more likely to
make significant contributions, they are
obviously not alone in making them.
After all, scientists do not begin by
king eminent (though the careen of
men such as MGssbauer and Watson
may sometimes give us that mistaken
impression).
The history of science
abounds in instances of basic papers’
having been written by comparatively
unknown
scientists. only to be neglected for years. Consider the case of
Waterston, whose classic paper on moItcuiar velocity was rejected by the
Royal Society as “nothing but nonsense”; or of Mended, who, deeply disappointed by the lack of response to
his historic papers on heredity, refUsed
to publish the results of his further
research: or of Fourier, whose ckmic
paper on the propagation of heat had
to wait 13 years before being finally
published by the French Academy (43).
Barber (44) 91as noted how the slight
professiorial standing of certain scientists has on occasion led to some of
their work, later acknowledged as significant, being refused publication altogether. And, correlatively,
an experienc&g
brd
Rayleigh’s (45) provides
an example in which an appraisal of
a paper was reversed once its eminent
authorship became known. Rayleigh’s
name “was either omitted or accidentally detached [from a manuscript], and the
Committee [of the British Association
for the Advancement
of Science]
‘turned it down’ as the work of one
of those curious persons called paradoxers. However, when the authorship
was discovered, the paper was found
to have merits after all.”
When the Matthew effect is thus
transformed into an idol of authority,
it violates the norx!l Of UIliversaliam
embodied
jn the institution
of Science
.
of knawiand curbs the advancement
edge. But next to nothing is known
about the frequency with which these
practices are adopted by the editors
and referees of scientific journals and
by other gatekeepers of science, This
aspect of the workings of the institution of science remains largely a matter of anecdote and heavily motivated
gossip.
The Matthew Efbct and Allocation
of Scicat&
Resoumu
One institutional version of the Matthew effect, apart from its role in the
reward and communication
systems of
science, requires at least short review.
This is expressed in the principle of
cumulative advantage that operates in
many
systems of social stratification
to produce the same result: the rich
get richer at a rate that makes the
poor become relatively poorer (46).
Thus, centers of demonstrated
scientific excellence are allocated far larger
rts~urccs
for investigation than centers
which have yet to make their mark
(47). In turn, their prestige attracts a
disproportionate
share of the truly
promising graduate students (48). This
disparity
is found to be especially
marked at the extremes (49): six universities (Harvard, Berkeley, Columbia,
Princeton, California Institute of Technoiogy, and Chicago) which produced
22 percent of the doctorates in the
physical and biological sciences produced fully 69 percent of the Ph.D.‘s
who later became Nobel laureates.
Moreover, the 12 leading universities
manage to identify early, and to retain
on their faculties, thae scientists of
exceptional talent: they keep 70 percent of the future laureates in comparison with only 28 percent of the
other Ph.D.‘s they have trained. And
finally, “the top twelve [universities]
are much more apt to reckit futurt
laureates who received degrees fern
other American universities than they
are other recipients of the doctorate;
half the iaureates who were trained outside the top twelve and who worked
in a university
moved into the top
twelve but only six percent of the
sampie of doctoral recipients did SO.”
These social processes of social selection that deepen the concentration
of
top scientific talent create extreme difficulties for any efforts to counteract
the institutional
consequences
of the
Matthew principle in order to produce
new centers of scientific excellence.
SThis account of tbt Ma&cw tiect
is another smaII exercise ia the ps~cb!Bociologicai anaiyh of the workings of science as a so&i in&Ution.
The initial problem is transformed by
-a shift in theepenpectin,
As
originally identi&d, the Matthew C&U
tnrw construed in terms of enhmxamf
of the position of already rfnjncnt
scientists who are g&n
Mrtionate credit in cases *of coiiaixmhm
or of independent multiple &WV&U.
Its significance
was thu8 cur&cd
to its implications
for the. system of scicsa#, By shiffing the angle
of vision, we note other me
w
of amsquences,
this time for t@ mmmunication system of S&XX. The Matthew effect may serve to m
the
visibility of contribtiom
to s&n~~ by
scientists of acknowkdgod S~JE&QJ a\d
to redUcc! the visibility of colltribotjo83
by authors who arc less well m.
We examine the psycbosocirl
um&
tions and me&misms
Und&y&this
effect and find a correlation betaracn
the redundancy
fun&on
of mo)tiple
discoveries and the focalizing m
of eminent men of scicocc-a
m
which is reinforced by the great vah~
these men place upon finding bsic
problems and by their self-assumm.
This seif-assUrance, which b +y
b
herent, partly the rutrlt of expcrimocs
and associations in -tie
&a&&
eIIyironments,
and partly a &t
of
later social validation of their gmihun.
encourages them to 8mmh a
l+SQ
but important problems and to hi+
light the results of their inquiry.
A
ma-1
version of the Me
principle is apparently invohred in those
processes of sociai selection that currently lead to the concentration
of scientific resources and talent (SO).
1.
2.
3.
Tbeambodaofobtrdntn~~~
of tkk
subH. A. Zuchmmn,
ufrivenity,
1965.
A-.
-1.
Rev: .a 633
intewiewr and the
-
are deadbed
r)lcdr,
calumM8
R. K. w
09s7).
B. 0. G-9
PmjemionaI cali& 1964h
Oe
chrrran
in
m--wkP
SC-
TIbb
D. Crlaa, Am-. ‘soclol. Rev.
w.
0.
3% 699 (196s’).
H8#%treln, Thr Sclrrrttpc co-may
(B&c Boo& Ncr Yo*’ l%r), chap. L
N. W. Storer. l%e Soefaf System ot S(He Rimbarr and Wimtaa, New York, 1966).
p. 106: se8 rl- ibid, pp. -26,
103406.
theai& c&albi8
U&
Ey %i!Y-t8. S. kok ;rd J. R C&e, Amer. Sodol. Rn.
3% 377 v9m.
9. I bare 8doptcd this term tat the m
phenomenon
from the tnonogr8p h on .tk
F-cb
Arrhmr by Arsene Hw,
H-8
du
41-9
FawuU
de l’A_
F-e
iPl?iSs,
1886).
10, This partial list of men who b8va &lb8
7.
. .
votk
‘ram
11.
of
“prize-winning
calibre”
it
derived
and His
Prittr
(Ellevier, London. 1962). an official publication
,f
the Nokl
prize-grantmg
academy
and
nstitutc, Nobclstlftelscn.
indebted
to Marshall
Childr
for sugpting
that this term,
inttoduccd
into
ccocomics
by James
S. Ducscnbcrry
in quite
~nothet
connection,
could
aptly
refer
to this
pattern in the cumulation
of prestige for suc:cuive
accomplishments.
For its USC in cco9Ofl¶iCS, see
Duesenbcrry,
Income,
Savings,
and
of Consumer
Behavior
(Harrard Univ.
Press, Cambridge,
Mass.,
1949).
sp. 11416.
Ihi,
process
of
a socfufly
rrintorced
rise
In ,rrpirat ions. as distinct from
Durkhcim’s
=owpt
of the
“insatiability
of wanta.”
is
:xamined
by R. K. Mcrton
in Anomie
und
Devkrnr
Behavior,
M.
Cllnard.
Ed.
(Fret
Pms, New York. 1964). pp. 213-242.
t.
Parsons,
The
Social
System
(Free
Press,
New
York, 19511, p. 127.
Max Webcr touches upon the convertibility
of position
in distinct
systems of stratification
In his classic
essay
‘“ClStatus, Party”
[Fmm
Max
We&w:
Essays
in
Socfofogy,
H; H. Gcrth and C. Wright Mills. Eds. (Oxford Univ.
Press, New York, 1916)).
Ihe
laureates are not rlone in notins thai
Nobel:
The
I am
26.
rhc Theory
12.
13;
14.
13.
i
prominent scientists tend to get the lion’s
rhulb of credit: similar observ8tions
were
made by less eminent scientists in the samo
pie studied
by Hagstrom
(set 3, pp. 24, 25).
cw
can be infcrrcd from
the pro16. A third
tocols
of interviews,
in which
the view
is
stated that,
had a paper
wtlttcn
by a com-
dlicatcd that they were familiar
with the work
bf a designated
list
of
120 physicists.
The
sltudy includes
checks
on Ihe validity
of
hcse visibility
scores.
>. J. dtSolla
Price has noted that “all crude
I neasurcs,
however
arrived
at.
show
10 a
approximation
incrcascs
f int
that
sclcncc
f :xponentially,
at
a
compound
interest
of
1rbout
7 per cent per annum,
thus doubling
10-15 years, growing
by a factor
i n size every
and by somcthmg
c,f 10 every lhalf~cntury,
1,ike a factor of a million
in the 300 years
which separate
us from
the seventccnthscn,ury invention
of the scientific
paper
when
:he process
began” [Nature 206, 233 (1965).
pp. 233-238).
B. Glass. Science 121, 583 (1955).
k.
for example.
H.
Men&.
In Communicordon: Conceprr and Perspecwes.
L. Thaycr,
(Spartan
Books.
Washington.
D.C..
Amer. Psychoforrfsr
:&)
pp 279-2950
21. ob9 il966).
See also ‘S. Herncr
(Science
128. 9 (1958)).
who
notes that “one of the
prcrtest
uimulams
to the use of information
is familiarity
with
its source”:
S. Hcmcr.
fnd. Eng. Chem.
46, 228 ( 1954).
Future
rnvcstigations
will
rcquitc
more
detaikd data on the actual proccsscs of selecting scientific
papers
for
varying
kinds
of
“rcrding”
and “skimming.”
But
the
data
now available
arc at least suggestive.
C
Man
27.
28.
1
.
,-
29.
30. On the concept
of functional redundancy as
distinct from “wa~cful
duplication”
in scicntific rcscarch. see R. K. Merton,
Europearr
31.
parat:vcly unknown scientist been presented
instead by an eminent
scientist,
it
would
have had a better chance of being published
an3 of receiving respectful attention.
Sy*
temrtic *information about such casts is too
18.
19.
20.
21.
+
not an efficient way of doing things. .I think
that most of the time. in respect
to tescatch,
duplication
of effort is a good thing.
32
an,
politics,
the
practicing
profcssionr,
or
rcliaion
for the recognition
of talent,
whatever its socirl
origins?
H.
Zuckennan,
“Patterns
of namc~rdtrjnq
among authors of scientific papers:
a study of
-aI
symbolism
and its ambiguity,”
paper
rtad .&fore
the American
Sociological
~ssocition,
Augmt
1967.
Dr.
Zuckcrman
will
not demean
herrtlf to give thcsc practices
their prcdcstincd
tag,
but
I shall:
plainly,
these arc instanccx
of Nobelesse
oblllre.
8. Berekaon, Graduate
Educatton
IU
lfnlted
States (McGraw-Hill,
New York,
1960). p. 55.
D. J. dcSolia
Price,
tftrlr
Science.
f3& SC&
tncy (Columbia
Univ.
Press, New York,
1963).
This
nattern of social functions
and individdysfunctions
is at variance
with the vigor-
33.
34.
Later
the
and
33.
36.
requirements
of the social system are in
suffictcnt accord to be functional for both
R.
L.
Ackoff
and the social
and M. H.
system.
Halbert.
An
Over-
atlonr Research
Study
of the Sclrnrlfic AC~lrlly of Chembrt
(Cue
institute of TechaoAogy Operations Research
Group,
Clevb
land
2%
1958).
on
~s~chofory
Project
37.
more than 1300 American physicists who in-
this
discussion.
I
consider
the
dys-
questioning
of mere authority.
Here,
other
institutional
spheres.
the prob=
lem is one of accounting
for patterns
of
cotncidencc
and
discrepancy
bctwccn
social
norms
and actual
behavior.
H.
Zuckcrman,
A nrrr.
Aociol.
Rev.
32, 391
In
mm.
Germane
resuirr
in expcrimcntoi
psychology
show
that
preferences
for
riskier
work
but
more
significant
outcomes
arc related
both
to high
motivation
for
achievement
and to
a capacity
for accepting
delay in gratification.
See. for example,
W. Mischel,
J. Abnormuf
Sot.
Psychol. 62, 543 ( 196 I).
this extent.
they engage
in the kind
of
bebavior ascribed to physicists of the “perfectionist ‘*
who
have
been statlstically
identified
by
the Coles
(8
tho.sc who
publish
less than they might
but whose publicrtions
nevertheless
have
a
considerable
impact
on the field. as indicated
by citations.
It is significant
that this type of physicist
was accorded
more recognition
in the form
of awards for scientific work than any other
types (including
the “prolific”
and the “mass
producer” types ) .
S. Bcnzcr,
in PhaRe and rhc OriRfns
of Molecular
Biology,
1. Cairns.
G.
S. Stent.
J.
D. Watson, Eds. (Cold
Spring
Harbor
Labo-
38.
Biology, Cold Spring
p. 165. This Festschri/f
shows that Dclbriick
is one of
Icicntists who generally
exercise
this
kind
of dcmandirw
judgment
on the publication
of their own work and that of their associates.
39. For some observations
on the prophylaxis
for this disease. see R. K. Menon,
On the
Shoulders
of Giants
Harcourt.
Brace
and
World,
New
York.
1967).
pp. g3-_85.
40.
has been noted (G. Williams, Virus Hwrrem (Knopf,
NW
York.
1959)J that the early
conhdtncc
of SCientlStS
in the mcas1cs vatcinc was a “paradoxical
feedback
of (Enden’s)
own
scientific
insistence,
not on bclicvinp,
but on doubting.
His fellow s&ntists trust John Enden
not to go overboard
on anything.”
41. This remains
a moot
conclusion.
Hovland’*
experiments
with
laymen
have
shown
that
the
scanBe communicauons
conudcrcu
are
less biased
when
attributed
to sources
of
high rather
than low credibility
[C. I. Hovland, Anrer. Pwcholonfs~
14. Y (1959)). In an
earlier
study.
Hovland
and
his
associates
found thrrt. in the cast of fucwol communica[ions, cherc is “equally
good
kaminv
of
what
was said
regardless of the credibility
of the commumcator”
[C. 1. Hovtrnd.
1. 1.
Janis, H. H. Kcllcy,
Communkahn
and Ptrsuusdon
(Yale
Univ.
Press,
New
Haven.
Conn..
1953
271)).
42. For an analysis of the sclf4ulfWlnq prophecy. see R. K, Merton.
Anrkch
Rev.
1918.
So5% (Summer
1948). rcpfintcd in -.
cial Tlrwrr urrd Swiul Strtwrrrre (
Press,
New
York.
19371, pp. 421-436.
43. Set W. K. hlcrton
who cttcs the following: R. H. Murray. Science ond Sclenrirts
in
the
Ninereenrh
Cenrtrr.v
(Sheldon.
London,
1923).
pp. .146w.t48:
D.
1. Watmn.
Scientists urt Hwnutt
Watts.
London.
1938 ), pp.
58. 80: R. J. Strutt
(Boron
Rayleigh).
John
Williurrr
Slrtrrr.
Third
Bitron
Hcrdrigk ( Arnald.
London. 1924). pp. 169-171.
134. 596 f I961 ). rcprintcd
44. B. Barber. Scirncv
and W. Hirsch.
Ed%.. T/rr
Social+
in g.v 01 Scitncr
(Free
Press. New
York. 1962).
pp. 539-556.
hy Barber
(44) from
R. J. Strutt.
45. Quoted
John
Wifliunt
Slrrrtl.
Third
Buror,
Rcrylelrrh
Arnold.
Londotl.
1924).
46. Dcrck Price pcrccivcd this implication of the
Matthew principle (Ndrrrt 206, 233 (196511.
47. D. S. (irecnbcrg. Satrrr4u.v Rev. (4 Novembcr 19671, p. 62; R. B. Bar&r. In Tht Pollrics oj Rrsturch
Public
Affairs Press. Washington,
D.C., 1966). p. 63. notcr that “in
1962, 38 per cent of all federal support went to
just ten institutions
and 59 per cent to just
H. Orlans,
Tht E#rcrs
of Ftd25:* See altral
Pr~~rumx
ON Hkhtr
Education
( Brookings Institution.
Washington, D.C.,
1962).
Cart&r
tcports
that.
in
Allan
M.
48. Thus.
19-3,
86 percent
of
(regular) National
Science Foundatron
Fellows
and 82 percent
of Woodrow
Wilson
Fellows
free to choose
their
place of study elected
to study in one
or another of the 23 leading universities (as
rated in terms of the quality of their graduate faculties)
(A.
M.
Carttcr.
An
Assessin
Graduate
Educarfon
rnemt
of
Qt1aftiy
(American
Council
on Education.
Washington, D.C.,
19661. p. 1081.
information
on
49. For this and other detailed
the career patterns
of lrrureatcs, see H. Zuck-
those
Clearly
(
It
1. p.
Fret
(11,
(
(
(
ermrn (1, 32).
to come
SO. Chancing
this paper, Richard
To
type.
Sctenrific lntormarlon
Exchange
In
(American
Psychological
Association, Washington. D.C.,
1963 ), vol.
1.
S. Cole
and 1. R. Cole, “Vlsibillty
and the
structural
bases of observability
in science,”
paper prcscntcd
bcforc
the American
Sociolo@c8l Association,
Auaust
1967.
In the Colcs’g study (24).
the term rgfsfbfllty
scores refers to percentages in a sample of
In
functions arsoccatcd with these functions of
great men of science. Idols
of
the
cave
often continue to wield great influcncc even
though
the norms of science call for the sysas
tioru
individuals
470 (l%l).
tematic
untutored optimism unforgctt;bly
by Adam Smith. who
speaks
of
it
24.
( 1963 ).
social stratificrtion
involving differentials in
lifechances,
thcrc
remains
the qucstlon
of
the extent to which talent among individuals
in the deprived
strata has gone unrccogni;Pcd
and
undcvclopcd,
and
its
fruits lost to
tiety.
MO= spccificaliy, we have yet 10
d-vet
whether or not the channels of
mobility are equally open to talent in vrriow
lnstftution8l rc8lms. Dots contemporary
science rfford grerter or less opportunity
than
expressed
“8 barrnomous order
of nature,
under divine
gtince,
which
promoter the welfare of man
through the opcratlon
of hire individual
propa&ties.”
If only
were that simple. One
of the prime problems
for sociological theoty is that of identifying the special condiunder which men’s propcnrities
and the
23.
237
I think
that if there
ate diflcrcnr
groups
in
diffennt
Iaboratortcs
working
on the same
thing,
their
approach
is sufficiently
different
(lo incrcasc
the probability
of a successful
outcome).
On the whole.
this is il good thing
and
not something
that
should
be avoided
for the sake of efficiency.”
So far as I know.
no investigation
has yet
been carried
out on pnciscly
this question.
At best sug;ertive
is the pcriphcrnl evidence
that papers of Nobel laurcatcs-to-bc
were
cited 30 times more often in the 5 years
before
their authors were awarded the prize
than
were the papers of the average
author
appearing
in the Citation
index
dunng
the same period. See i. H. Shcr and E.
“New
t-Is
for improving
the effecGuAcid,
paper prescntcd
at the
tivencss
of research.”
2nd Conference on Rcscarch Program Effectivencss.
Washington.
D.C..
July
1965:
H.
Zuckcrman,
Scf. AtpIer.
217. 25 (1967).
R. K. Merton,
Proc. Amer. Phil. Sot. 10%
ous
22,
4.
the laureates questioned
the
ready
redundancy
of
assumption
that
rescrrch
effort
ncccssrrily
means
“wasteful
duplication”:
“One often hears, especially when large
amounts
of money are involved.
that duplication of effort should bc avoided.
that this is
spans
for detailed
study.
compensatory
pattern
can only
obtain,
17. This
of course,
among
scicntlsts
who
ultimarcly
achieve
recognition
with
its associated
further
rewards.
But,
as with
all
systems of
*
1. Socfol.
One
of
r;ltory of Quantitative
IHarbor, N.Y., 1966).
1 as
upon
the manuscript
of
1. Russ&
a mokcular
biologist
of more than passing aCuUaintanCc.
has informed
me that a well-known
textbook
in organic chemistry
(L.
F. Ficser and M.
lo
Oruanic
Chemistry
F jeser. In runfucrion
(Heath.
Boston.
1957 )] refers lo the “empirical rule due to Saytzeff
( 1873)
that in dehydralion of ;Ilcohols. hydrogen is eliminated
prefcrcntially
from the adjacent
carbon
atom
that
is poorer
in hydrogen.”
What
makes
the rule germane
to this discussion
is the ac“MA~Ew,
XXV,
footnote:
companying
but from him that bath not shall
29, ‘.
bc taken
away
even that which
hc hath.’ ”
Evjdcntly the Matthew
effect transcends
rhe
world of human behavior and s&al
process.
Earlier
versions
of this discussion
were prcscntcd
before
NIH
and AAAS.
The
work
summarized
was suppotted
in part
by NSF
grant
GS-960
to Columbia
University’s
pro,
gram in rhe sociology
of science. This 8rtiCle
No.
A-493
of the Bureau
of
is publication
Social Research, Columbia
UnivcnitY
. .
51.
Applied
g